PLM: between theory and practice

With PTC Windchill, you have all your up-to-date product data at your fingertips. At least, in theory. In practice, every engineer knows that it takes a little more to get information out of a system as extensive as a PLM database. Especially for those who don’t work in Windchill on a daily basis, it’s not easy to find the right information: clicking through complex screens, searching for the right BOM, losing time on manual lookups, and so on. 

But what if you could just ask for that data in plain English, just like we’re used to with something like Copilot? Imagine not having to click through a bunch of screens, but simply asking, ‘What’s the structure of this assembly?’ or ‘What’s changed since the last release?’ and getting an answer, or even better, having an action performed.

Making AI valuable for engineers

We see it often enough: companies are experimenting with AI, but struggle to get actual value out of their AI tool. This is also true for engineers: they feel the pressure to ‘do something with AI,’ but product data – think 3D designs, electrical schematics, stress calculations – is a lot more complex to unlock than, say, Excel or Word documents. Moreover, product data is at the core of an organisation: you don’t just want to plug it into any public AI model. To bridge that gap between engineering needs and AI promises, a bunch of our colleagues got together.

That’s how Savaco Windchill PLM Buddy was born: a secure AI layer on top of Windchill that makes your PLM data available in natural language, directly through Microsoft Copilot. You ask your question in plain language, Copilot pulls the answer from your PLM, and you can immediately turn it into action. No need to worry about security, governance, and control. Your data doesn’t leave your environment, and PLM Buddy works with the exact same access rights as your Windchill platform. Anyone who asks PLM Buddy a question will see exactly what they’re allowed to see in Windchill, and nothing more.

Learn more about Windchill PLM buddy
Watch Savaco Windchill PLM Buddy - your Digital Thread, now AI-native on YouTube.

What is Windchill PLM Buddy exactly?

Windchill PLM Buddy runs on MCP server technology, a way to make AI models ‘talk’ to different source systems. That’s exactly what PLM Buddy does: it pulls data from Windchill based on natural language questions, just like LLMs have been doing with the public internet for a while now. Windchill PLM Buddy comes with four standard functionalities that you can ask for directly from Copilot and immediately save time: 

  • Query BOMs and structures
  • Query information about parts, such as status or metadata
  • Create or update problem reports
  • Create change notices 

But it doesn’t stop there: Windchill PLM Buddy is not a finished off-the-shelf tool, but a building block. Together we will continue to build on it, in agile sprints, to add new functionalities that are precisely tailored to how your organisation works. This could be about having new actions performed in Windchill, but also about retrieving new data from other systems in your digital thread: think for example of ERP or QMS. So you start with a set of features that immediately save you time, but you also have a tool in hand that will continue to grow with your own needs. 

Watch On Track #12 | Haal meer uit je PLM-data met AI (Windchill PLM Buddy) on YouTube.

Why we’re proud of Windchill PLM Buddy

Windchill PLM Buddy was born when two areas of expertise within Savaco came together. It started with a concrete, everyday problem that we often see with our customers: navigating and making product data accessible in Windchill is not easy, no matter how powerful Windchill is. And we decided to work on a solution with different teams. On the one hand, our Data & AI team, who, with their knowledge of MCP server technology, data warehousing, and governance, can build AI agents that not only function, but are also safe and manageable. On the other hand, our Windchill experts, who have been working closely with industrial players and engineering teams for years. They know how a BOM is structured, how change processes work, and what engineers really need.  

An AI team without PLM knowledge builds something that is technically great but doesn’t help the engineer. A PLM team without AI knowledge remains stuck with the idea. By bringing both worlds together, PLM Buddy became something that works in practice. Lieven and Aster tell us exactly how we did it in our On Track podcast.

Listen on Spotify
Webinar

Meet Windchill PLM Buddy

Want to know what Windchill PLM Buddy can do for your team? Register for our webinar and get a first-hand look at the new way to query your PLM data.

Watch the webinar